Machine Learning learning notes

Notes to learn Machine Learning from zero to hero using Visual Notes by @carlogilmar. Hope this could be useful for all of you.

Table of contents
Supervised Learning - Week 1
S1: Machine Learning Basics
S2: Linear regression
S3: Cost function formula
S4: Cost function params
S5: Gradient Descent

Summarize of concepts

ML Summary Topics

1. Machine Learning

• What it is?
• Types of ML
  • Supervised Learning
    • Linear Regression
    • Regression Models
  • Unsupervised Learning

2. Linear Regression Supervised Learning

• Model: inputs (features) X that predict Y outputs(targets) using a training set.
• Linear regression function model with one variable
• Cost function: Evaluate the error in your model.
• Gradient Descent (average loss of the features): Minimize the cost function.
• Uncertainty
  • Epistemic Uncertainty: lack of knowledge in the model.
  • Aleatoric Uncertainty: lack of knowledge in the data.
  • Capturing Uncertainty data using a probability distribution.
• Maximum likehood estimation

3. Multiple Variables and Vectorization

S1 Machine Learning Basics

💡 Insights:

• Machine Learning refers to the computer's ability to learn without being programmed.
• There are types of ML: supervised and unsupervised learning, recommender systems, and reinforcement learning.
• Supervised Learning are algorithms that learning from X inputs to generate Y outputs.
  • There are two types of Supervised Learning algorithms: regression and classification.
  • Regression algorithms are for predict a number, there are infinite possible outputs.
  • Classification algorithms are for predicts categories.
• Unsupervised Learning works with data that is not associated to an specific output.
  • This algorithms try to find structure in the data.
  • Clustering is one kind of this algortihms.

S2 Linear Regression

S3 Cost function formula

S4 Cost function params

S5 Gradient Descent